For samples primarily under 75 years of age, some cardiovascular risk factors (CVRFs) have been associated with increased risk of cognitive decline, dementia, and Alzheimer's disease. Although frequently interpreted as applying generally to all elderly, findings from our group and others suggest that such associations are greatly diminished or even reversed in more elderly samples. If, as these results suggest, the effects of some CVRFs do not reflect stable cardiovascular risk across the life span, these effects may reflect enduring traits possessed by some protected survivors. Alternatively, they may reflect adaptive changes in some elderly as they progress to late old ages while maintaining good cognition. Aim 1 of this R21 will determine the applicability of each of these three models of successful cognitive aging (SCA; living at least 90 years free of cognitive impairment) to the relationships of cardiovascular risk factors (CVRFs) with neuropsychological phenotypes. We will examine the effects of CVRFs on neuropsychological phenotypes by comparing proband and relative groups highlighted by the respective SCA model. The project will be performed in Costa Rica, which has high life expectancy and reduced genetic and cultural heterogeneity. In Aim 2, the Aim 1 results will be used to design a multi-site R01 on the genetic and epigenetic basis of SCA, concentrating on genes affecting CVRFs associated with each model. Aim 3 is to build research capacity in Costa Rica upon an already existing infrastructure. We will train the experienced and dedicated Costa Rican investigators on this research team, so they can perform independent studies of SCA and aging in Costa Rica based on data from this R21, the R01, and other collaborations.